# Scalable and Performant Data Loading

Source: [Scalable and Performant Data Loading](<https://www.sota2.com/research/paper/68fa78411e209131cea7774b>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-12T05:00:12.407Z

Authors: Moto Hira, Christian Puhrsch, Valentin Andrei, Roman Malinovskyy, Gael Le Lan, Abhinandan Krishnan, Joseph Cummings, Victor Bourgin, Olga Gerasimova, Miguel Martin, Gokul Gunasekaran, Yuta Inoue, Alex J Turner, Raghuraman Krishnamoorthi

Published: 2025-04-23T19:59:43.000Z

## Abstract

We present SPDL (Scalable and Performant Data Loading), an open-source, framework-agnostic library designed for efficiently loading array data to GPU. Data loading is often a bottleneck in AI applications, and is challenging to optimize because it requires coordination of network calls, CPU-bound tasks, and GPU device transfer. On top of that, Python's GIL (Global Interpreter Lock) makes it difficult to gain performance improvement from multi-threading. We found that when data preprocessing functions release the GIL entirely, it is possible to execute them concurrently in a thread pool, thereby improving the workflow performance. Our benchmark shows that compared to the PyTorch DataLoader, SPDL can iterate through the ImageNet dataset 74% faster while using 38% less CPU and 50GB less memory. When training ViT-B/16 model, SPDL can send data to the GPU at a speed that does not starve the training. Additionally, when using SPDL on Python 3.13t, without changing any code, the throughput is further by improved by 33%, thanks to the disabled GIL. SPDL can improve the performance of current AI model training, and receives further performance improvements when Free-Threaded Python is adopted in production systems. SPDL is available at https://github.com/facebookresearch/spdl.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2504.20067>)

[PDF](<https://arxiv.org/pdf/2504.20067.pdf>)
